QuestionsCan a machine think?The machine originates nothing: it carries out what it is ordered
Can a machine think?Understanding is missing

The machine originates nothing: it carries out what it is ordered

Lady Lovelace's objection

Ada Lovelace 1815 · 1852She described Babbage's Analytical Engine, saw that it could handle more than numbers, and set the limit it does not cross: it originates nothing.
See profile
What it claims

The Analytical Engine creates nothing: whatever it produces comes from what it was ordered to do.

Why

The machine only does what we know how to order it to do and anticipates nothing, and creating takes more than that.

The same idea through history
Charles Babbage1837Babbage designs the Analytical Engine: a punched-card machine with a store and a mill, which can be programmed for very varied calculations.
Ada Lovelace1843Translating an article by Menabrea on this machine, Lovelace adds her own notes. In Note G she warns against exaggerated ideas: the machine only does what we know how to order it to do.
In short

"The machine originates nothing: it carries out what it is ordered" is an argument by Ada Lovelace (Notes on Menabrea, Note G, 1843, 1843) on the question "Can a machine think?" (answer: understanding is missing). Its claim: The Analytical Engine creates nothing: whatever it produces comes from what it was ordered to do. According to AskPhilosophy's logic check (Z3 solver), the argument is valid only if one grants a premise the author leaves unstated (P3). AskPhilosophy presents 5 objections, 0 of them with a documented reply, and the original text behind each premise, checked word for word.

Notes on Menabrea, Note G, 1843
Valid with an implicit premise

In 1843 Ada Lovelace describes Babbage's Analytical Engine, a mechanical calculator programmed with punched cards. She sees very far (the machine could handle more than numbers), but she sets a limit: one must not credit it with powers it lacks. The machine carries out what we know how to order it. It creates nothing. This sentence became "Lady Lovelace's objection": a machine does not really think, because it has no initiative and no understanding of its own of what it does.

P1The Analytical Engine has no pretensions to originate anything. It can do whatever we know how to order it to perform.
P2It can follow analysis, but it has no power of anticipating any analytical relations or truths.
P3 · implicit premiseTo create is to produce something that was fixed beforehand neither by the rules nor by the one who gives them. A machine that does only what it is ordered, anticipating nothing, therefore creates nothing.
So the Analytical Engine creates nothing: whatever it produces comes from what it was ordered to do.

click a premise to reject it

If you accept every premise, logic forces you to accept the conclusion. So the debate is about the premises.
Checked by a logic solver: the premises cannot all be true while the conclusion is false.
Without the implicit premise: Without P3, the computer finds a case where the machine does only what it is ordered and anticipates nothing, and yet it creates: a pupil who applies the rules of grammar and verse to the letter writes a poem that nobody, neither he nor his teacher, had foreseen. Following rules does not prevent producing something new.
The argument is valid once P3 is added. P1 and P2 say what the machine does (carry out, follow) and what it does not do (anticipate). P3 says what "create" means. Lovelace does not state it as a premise: she implies it by writing "originate". The whole debate is about P3 and about the word "order" in P1: if ordering the machine does not mean knowing what it will produce, then a machine can surprise us, and perhaps create.
Analogy. A piano plays only what the pianist makes it play: we do not say the piano invents the music. For Lovelace, the Analytical Engine is an extraordinarily powerful instrument, but an instrument: the invention comes from the one who programs it.

Objections

5

Each objection is a conversation: click to see who replies and where the debate stands.

Alan Turing 1950 · attacks premise 3

He asks who can be certain that "original work" he has done was not simply the growth of a seed planted in him by teaching, or the effect of following well-known general principles. If that is the criterion, humans may not create from nothing either, and the definition of "create" in P3 is too strong.

no reply
see the argument
No reply possible: objection raised in 1950, after the author's death (1852).
Where the debate stands (our summary)Open point. Turing challenges the idea that a human creates "from nothing". Lovelace's defenders can answer that, even if a human is shaped by education, he chooses what pleases him, what is false or beautiful, which the machine does not do by itself. Neither reading has prevailed (our summary).
Alan Turing 1950 · attacks premise 1

Machines take him by surprise very often. Not because they do something other than what they were ordered, but because he does not work out in advance what they will do. One can therefore order a machine to run a program without knowing what it will produce: "knowing how to order it" does not mean "knowing what will come out".

no reply
see the argument
No reply possible: objection raised in 1950, after the author's death (1852).
Where the debate stands (our summary)Disputed. Turing shows that "ordering" and "foreseeing" are two things; his critics answer that being surprised by a result does not suffice to say the machine invented it: the surprise is in the mind of the observer. Turing knows this and says so himself in the same passage (our summary).
Alan Turing 1950 · attacks premise 2

If a machine learns, the one who teaches it is often very largely ignorant of what goes on inside, although he may still predict his pupil's behavior to some extent. In contrast to an ordinary computation, where one follows each step, a learning machine therefore ends up doing things that nobody programmed one by one.

no reply
see the argument
No reply possible: objection raised in 1950, after the author's death (1852).
Where the debate stands (our summary)Open point. Turing describes a machine whose behavior is not written line by line. The question is whether that is enough to answer the objection: one can say that the programmer did order everything (the learning method), or that the machine now has a form of initiative. No consensus (our summary).
D
Douglas Hartree 1949 · attacks the conclusion

Hartree quotes Lovelace's sentence, then specifies that it does not rule out, in principle, building one day machines that "think for themselves", for instance by giving them conditioned reflexes that would serve as a basis for "learning". But the machines built or planned at the time did not have this property. Lovelace's limit therefore describes the state of machines, not all possible machines.

no reply
No reply possible: objection raised in 1949, after the author's death (1852).
Where the debate stands (our summary)Hartree narrows the scope of Lovelace's sentence to the technology of her day. What remains is more modest: a machine that merely executes originates nothing, but nothing says a machine will never do more. Turing shares this view (our summary).
M
Margaret Boden 1990 · attacks premise 3

Boden refuses to treat creativity as a mystery: she describes it as the ability to combine familiar ideas in unexpected ways, to explore a space of possibilities defined by rules, and sometimes to transform those rules themselves. Part of these operations can be described by programs. The question therefore becomes: which form of creativity can a machine have, and which escapes it?

no reply
Lovelace never answered this objection.
Where the debate stands (our summary)Disputed. Boden holds that "creating" is not incompatible with "following rules": P3 defines creativity badly. Her critics ask whether a machine exploring a given space of rules understands the value of what it finds, or whether it is still the human who recognises it (our summary).